With rigorous economic research and practical policy solutions, we focus on the issues and institutions that are critical to global development. Explore our core themes and topics to learn more about our work.
In timely and incisive analysis, our experts parse the latest development news and devise practical solutions to new and emerging challenges. Our events convene the top thinkers and doers in global development.
Global poverty is decreasing, but billions of people still do not have the resources they need to survive and thrive. Economic growth can reduce poverty, but it can also drive inequality that generates social and economic problems. And efforts at domestic resource mobilization through taxation, though critical to funding the SDGs, can negatively impact the poor. In this work, CGD experts offer suggestions to improve how the world tracks and tackles poverty and the inequities the international global system creates.
Get Poverty, Inequality, and Growth Updates
CGD experts offer timely research, analysis, and policy ideas for the world’s emerging development challenges. Sign up to get the latest updates from CGD!
Recently, the World Bank published its latest Global Economic Prospects report (GEP), which highlights a welcomed cyclical recovery for all major regions of the world following recent slow growth. I was pleased to participate in a panel discussion at CGD analyzing the report’s findings, and to share my perspectives both on its implications and on future global outlooks—especially for emerging market and developing economies (EMDEs).
Although short-term growth prospects have improved for EMDEs, with an expected average of 4.5 percent for 2018 and 4.7 percent for 2019 and 2020, the report stresses that this is no time for complacency. Despite improving economic activity, potential long-term growth is predicted to extend its decade-long falling trend in these economies. Undoubtedly, this impedes the likelihood of any foreseeable convergence of per capita income with those of advanced economies. So then, what should policymakers in EMDEs do? The answer is straightforward: they should implement proper reforms to improve potential growth. As the GEP report emphasizes, the right policies should aim to boost human and physical capital and improve productivity through reforms in infrastructure, education, health systems, labor markets, governance, and business climate.
The question then becomes, how likely is it that these necessary reforms—to improve productivity and convergence prospects—will materialize? The answer depends on two factors: the capacity and the willingness of policymakers to pursue proper reforms.
Many reforms inevitably require large financing capacities. This is often the case, for instance, with infrastructure investments. Yet, one does not need to look further than the IMF’s latest World Economic Outlook to recognize the vastly limited fiscal space amongst EMDEs, especially commodity exporters. Consequently, policymakers from many EMDEs are taking advantage of the currently favorable conditions in international capital markets, which continue to offer attractively low interest rates. This outcome is supported by little risk differentiation among EMDEs’ debt by investors, as reflected by very low spreads (the difference in yield between EMDE bonds and US Treasury bonds of a similar maturity). This points to an underpricing of risks.
The Latin American Committee on Macroeconomic and Financial Issues (CLAAF), over which I preside, explored this underpriced risk phenomenon in our most recent statement. We noted the resounding success that EMDE nations—even those with low credit ratings—have had in issuing international bonds. This was the case both with Côte d’Ivoire, which successfully sold bonds internationally during a military riot, and with Iraq, which issued a bond with demand that far exceeded what was offered, to name a few. This growing debt accumulation is alarming as there are many external risk factors (as noted in the GEP report) that can increase both the costs of servicing this debt and the likelihood of defaults. Therefore, for the rise in debt levels to be justified, they must be accompanied by a willingness on the part of EMDEs’ policymakers to pursue reforms that will lead to future growth—a necessity to amortize these liabilities and avoid potential crises.
Although there are numerous reforms where funding is not the binding constraint—labor market reform being a point in case—all reforms are bound by policymakers’ willingness to pursue them. This is the part of the story that meets the reality of the current political economy in many EMDEs.
In many EMDEs, there is growing resentment towards democracy. Recent data from Afrobarometer depicts a slowdown in the demand for democracy in Africa since 2008, with 24 of the 36 countries surveyed experiencing negative or neutral changes. Similarly, surveys by the Latin American Public Opinion Project at Vanderbilt University show that support for democracy in Latin America has dwindled from 66.4 percent in 2014 to 57.8 percent in 2016/17. This trend is supported by a frustrated new middle class in EMDEs, who experienced a significant increase in their standards of living until recent years. According to a poll by Latinobarómetro, 37 percent of people in Latin America self-identified as belonging to the middle class in 2011, and this value rose to 42 percent in 2017. Surely this middle class feels more entitled as they have seen solid gains in their consumption and societal status, which they wish to continue experiencing and will not be willing to give up. A similar picture can be painted for the middle class in many other EMDEs.
As noted by Daniel Zovatto, in Latin America, this middle class discontent is materializing as we approach a marathon of important presidential elections. This heavy election cycle appears to be a common occurrence amongst EMDEs. For example, there are 12 Latin American presidential elections in the next two years and 14 African elections (general or parliamentary) in 2018 alone. This creates a good moment to propel populist political candidates to power who promise to secure the improved lifestyles of the middle class, which they strongly fear losing. If this were to materialize, the easily accessible external funding available to policymakers would likely be used to meet the demands of this vulnerable middle class through short-term benefits—not long-term productivity investments.
A great risk facing EMDEs is that although external funding conditions are favorable, the current domestic political economy suppresses the incentives for proper reform. The outcome then could be a major debt problem. Let’s hope that citizens of EMDEs choose the right leadership to put in place the proper reforms to increase the much-needed potential growth.
The World Inequality Report updates and extends the famous elephant curve, showing slower gains for much of the globe, and even more concentration of economic growth in the top 1 percent.
The "elephant graph" created by Christoph Lakner and Branko Milanovic in a 2013 World Bank paper has become perhaps the most talked-about chart in international economics over the past few years. The simple graph shows income gains at each point of the global income distribution for the 20 years spanning the fall of the Berlin wall to the 2008 financial crisis. China and India's rapid growth constitute the bulk of the elephant, driving down global inequality, while the global 1 percent lift the elephant's trunk, and in between the losers from globalization saw zero income growth—a feature frequently cited by pundits to explain the advent of Trump and Brexit.
The elephant graph is almost five years old now, so it's time for a refresh. Brace yourself. It's not good news.
In December, the team of economists from Berkeley and the Paris School of Economics who have pioneered the study of top incomes—including Facundo Alvaredo, Lucas Chancel, Thomas Piketty, Emmanuel Saez, and Gabriel Zucman—launched the first World Inequality Report (WIR). Compiling all existing household survey sources with their own research on top incomes, they were able to extend the time period covered, and found a somewhat different picture:
Figure 1. The elephant curve of global inequality and growth, 1980-2016
The elephant's trunk is getting really, really long. As the caption notes, the top 1 percent captured 27 percent of total income growth from 1980 to 2016. The longer neck reflects the use of more and newer data on top incomes from tax records, which is the specialty of the World Inequality Report team. In contrast, the original Lakner and Milanovic graph was based on household survey data compiled by the World Bank, which the authors readily acknowledged are a poor measure of high incomes—simply because sample surveys are unlikely to randomly capture the (very few) super rich.
The explosion of top incomes at the far right of the graph now dwarfs the whole picture—leading my colleague Charles Kenny to helpfully suggest this looks more like a brontosaurus, or alternatively "Nessie," Disney's version of the Loch Ness monster, than an elephant.
Maybe it's a Trumpist elephant after all. The WIR authors explicitly label the trough of low growth with the bottom 90 percent in the United States and Western Europe—a claim that has previously been challenged. Adam Corlett of The Resolution Foundation (ht Caroline Freund's blog post here) argued the trough in the Lakner-Milanovic graph was driven more by Japan and Eastern Europe than Trump or Brexit voters. In reply, Lakner and Milanovic noted that the basic ranking of higher growth for the Asian middle class and the global elite, and lower growth for the middle class in the rich world is robust to any version of the calculation—and that seems to remain true in the new WIR version.
The whole elephant is sinking. In the original graph, the body of the elephant peaked at about the global median income, which saw roughly 75 percent cumulative income growth over 20 years, equivalent to an annual rate of just under 3 percent per annum. In the new Alvaredo et al version, the highest growth rate apart from the super rich comes for people at about the 20th percentile, who saw a 120 percent cumulative increase over 36 years—which is about half a percentage point slower per year.
Is that last discrepancy a function of the different time periods, or something else? It seems to be something else. I downloaded the replication files from the World Inequality Report, made sure I could perfectly reproduce the main elephant curve above, and then simply changed the start and end dates to limit them to 1988 to 2008 a la Lakner and Milanovic (instead of 1980 to 2016 in the new version). The result is still wildly different from the original elephant.
Figure 2. The original elephant graph
Source: Freund, Caroline. PIIE. Data courtesy of Lakner and Milanovic. World Bank. 2013.
Figure 3. The new elephant graph from the World Inequality Report looks more like the Loch Ness monster.
These are very different animals. The elephant has drowned! To be honest, I'm really not sure what's going on.
I suspect much of this discrepancy has to do with methodological differences, particularly in how the WIR team has tried to reconcile survey estimates to economic growth rates from national accounts. But the discrepancies are big and probably demand further explanation, given the inordinate popular attention that the original elephant graph received. A couple basic facts remain robust though: after a period of growing global inequality in the mid-20th century, incomes around the world are now converging with the rise of large developing countries, and—lest we feel too complacent—that equalizing force is partially offset by breakaway growth for a tiny global elite.
Thanks to Facundo Alvaredo and Branko Milanovic for helpful clarifications by email. And kudos to both research teams for posting replication files online. All the interpretations and any errors here in the calculations are obviously mine alone.
The World Bank’s Global Economic Prospects 2018 recognizes that the global economy is enjoying a long-awaited broad-based cyclical recovery. In this favorable environment, the Bank expects growth in emerging and developing countries to continue during the next couple of years. But this is no time for complacency. Forces depressing potential output growth will continue unless countered by structural policies. While most commentators focus on the recent cyclical upturn, the new World Bank report presents a sober analysis of long-term growth prospects. Director of the World Bank's Development Prospects Group, Ayhan Kose will give a brief presentation of the report and will then participate in the panel discussion, moderated by CGD president, Masood Ahmed.
What are the challenges and opportunities for growth in the Middle East, North Africa, Afghanistan and Pakistan (MENAP) region? In his presentation, Jihad Azour will present the IMF’s latest economic outlook for the MENAP region. He will argue that growth has not been fast enough and has not created sufficient opportunities to address high levels of unemployment. The presentation will be followed by a panel discussion on the main impediments to growth and highlight the policy priorities to durably increase it and make it more inclusive.
As Lant Pritchett reports, the World Bank has introduced two new poverty lines: $3.20 for lower middle income countries, and $5.50 for upper middle income countries. I’m with Lant that this is broadly a good thing. But the process by which the World Bank came up with its new poverty lines suggests it might be worth revisiting some of the pitfalls of income thresholds at the individual or national level. Grouping people’s or countries’ income classifications can help illustrate trends and features—but woe betide the analyst who takes the classifications too seriously and acts as if they reflect discontinuities in the underlying income distributions. They don’t.
The source of the new Bank poverty lines is illustrated by the below chart from the Bank’s blog announcing the new poverty cutoffs. $3.20 is the median poverty line for lower middle income countries (GNI/capita between $1,005 and $3,955), while $5.50 is the median poverty line for upper middle income countries (GNI/capita between $3,956 and $12.236). This has some odd implications. Imagine a country sees distribution-neutral economic growth that takes it from a GNI per capita of $3,000 to $4,000—crossing the threshold into upper middle income status. The average person’s income will go up by about 25 percent. The average poor person’s income will go up about 25 percent. But because the poverty line has changed from $3.20 to $5.50 (an increase of 72 percent), the number of people that the Bank says are poor will go up, too.
Figure 1. National poverty lines increase with national income
Figure by Francisco Ferreira and Carolina Sanchez, World Bank (Click to view)
Or put it another way: “You are 'poor' because you live in 1 of 56 countries with a GDP per capita between $3,956 and $12,235 on less than the median poverty line of those 56 countries” is not something that is immediately intuitive. And one big reason it isn’t immediately intuitive is that the World Bank’s income thresholds are utterly arbitrary lines drawn through per capita income data, which means a poverty line set at the median of those countries is pretty arbitrary, too.
Grouping people and countries by income categories even if they are just selected for convenience can be helpfully illustrative. That the proportion of people worldwide living on less than $1.90 a day has dropped from somewhere around half some time the 1950’s to around 10 percent today tells you something about declining absolute deprivation and rising incomes. That the number of countries where GNI per capita is below $1,005 (AKA the middle income threshold) has been rapidly declining suggests something similar. But any such line has to be used with care, otherwise it can obfuscate more than it illuminates. I’d argue that’s the case with the World Bank’s new poverty lines.
To take other examples of when such division leads to lack of clarity:
People can claim “global poverty is getting worse” because the absolute number of people worldwide living on less than (say) $7.40 a day has been rising.
Think instead: The number has been increasing because people in poor countries are living longer (so there are more of them), not because they’ve been getting poorer. The income of the average person living under $7.40 a day has been rising.
A bunch of countries get a little bit richer and cross the line from low to middle income and suddenly the “global poverty problem” has changed and is about middle income countries.
Think instead: Countries with a GDP per capita just a bit above $1,005 are still really poor, and the poorest people worldwide are still concentrated in really poor countries.
China has pretty much wiped out $1.90 poverty, is at the upper end of upper-middle income status and so you might think it is now rich.
Think instead: average consumption in the country is still only $288 a month and 400,000,000 people (more than the population of the US) still live on less than $5 a day (give or take, one third of the US poverty line).
People cross the threshold of $1.90 or $5 or $10 a day and become valuable middle class citizens who will preserve liberal democracy and improve the quality of government services.
Think instead: some attitudes vary a lot by absolute income, others don’t, but there’s no evidence of uniquely middle class values emerging at any particular income cutoff.
That few countries (in the 1990’s) went from low to middle income or (in the last few decades) have gone from middle to high income, suggests that countries a bit below one of these arbitrary lines have trouble crossing that line—escaping a poverty trap, or the middle income trap.
Think instead: growth varies a lot across countries and over time, but there’s no evidence (see this CGD blog post and working paper) of growth slowing when countries near a GNI per capita of $1,005 or $12,235.
I’m not citing those beliefs because most of the original authors were better scholars than to suggest such things in nearly so stark terms, but I do think these interpretations are pretty common in the development discourse. And they’re driven by over-interpreting income cutoffs. In short, existing poverty and income classification lines obscure an underlying truth—there is no discontinuity at the income line except that created by the existence of the line itself.
The announcement via this blog that the World Bank will now routinely calculate and report multiple international poverty lines is an important advance for development.
Perhaps the big debate in development today is between on one side those who want to “define development down” to a set of low bar, individualized targets, attainable with logistical programs (like the MDGs), and those who want to retain the original vision that development was fundamentally a long-run national and social process of attaining economies with high productivity, polities accountable and responsive to their citizens, administrations and organizations capable of achieving implementation-intensive targets, and a society based on fair treatment of all (like a set of Millennium Development Ideals).
The exclusive reporting of international poverty based on the penurious “dollar a day” poverty line (morphed by inflation into the less lyrical “$1.90 a day” line) was a tool for “defining development down.” The only advantage of the “dollar a day” line was that it was as low as a poverty line could reasonably be (in fact this was its intellectual justification—that no country had lower lines so this was the lowest possible line). Someone below “dollar a day” was for sure poor and hence no one could accuse the “dollar a day” standard of overstating poverty.
But the “dollar a day” standard never had, nor was ever meant to have, any rationale or credibility as a standard for who was “not poor.” If poverty is regarded as “an unacceptable deprivation in human well-being” then there were billions of people who were above the “dollar a day” poverty line but were poor by many legitimate definitions and standards of poverty. What is “unacceptable” deprivation is scaled by norms of what is possible and hence country poverty lines increase as average country income increases. This is also reflected in other approaches to poverty like multi-dimensional poverty and in participatory approaches to poverty, neither of which produce anything like “dollar a day” as the univocal or a consensus definition of poverty.
The conceptual weakness of the low bar poverty lines like “dollar a day” (and its twice as large sibling “two dollars a day”) is revealed in the ridiculousness (as pointed out by Nancy Birdsall and others) of dividing the global population into “poor” “middle class” and “rich” and referring to people just above “two dollars a day” as “middle class.” A low bar poverty line necessarily treats both people just above the poverty line and people in considerable comfort and prosperity as “not poor” and hence necessarily creates false equivalences in which those just above and just below a poverty line were considered to be different (one “poor” and one “not poor”) but two people above the poverty line with very different incomes were treated the same (both “not poor”).
The problem with defining the development agenda down to measures like “eliminating extreme poverty” and other “kinky” goals (as Todd Moss points out this extends to low bar goals in other domains like energy) is that is excludes most of the population in most developing countries as relevant for the “development” agenda. Indonesia, for instance, is a lower middle income country facing many challenges and no one considers Indonesia “fully developed” on any definition. Yet in 2016 headcount “dollar a day” poverty was only 10.6 percent so by this standard 9 in 10 Indonesians were not included in an “eliminate extreme poverty” development agenda.
The World Bank’s move away from the exclusive reporting of a low bar poverty line to the recognition of the legitimacy and importance of multiple poverty lines is an important step forward to recognizing development as a broad and inclusive agenda.
In 1995, India’s Self-Employed Women’s Association (SEWA) organized women waste pickers in Ahmedabad into a cooperative to improve their working conditions and livelihoods. Over time, this informal arrangement evolved into Gitanjali—a women-owned and -run social enterprise. With support from key partners, Gitanjali has generated social value, providing its members with safe and dignified work while increasing their earnings. While Gitanjali faces challenges in becoming a fully self-sufficient social enterprise, its experience offers insights for other initiatives seeking to provide opportunities for women to transition from informal to formal work.